PROJECT SUMMARY The concept of cancer precision medicine is simple: patient treatments are guided by the molecular features of their tumor. In practice, however, the situation is more complex. Only a minority of patients at present benefit from the approach. A number of scientific challenges must be overcome before cancer precision medicine can become a reality for all patients. These challenges are the focus on the present research proposal. For example, we now know that gene expression, not DNA-level aberrations, are the most predictive of drug response, and yet the entire clinical genomic testing enterprise is focused on DNA. We therefore need to develop methods suitable for quantitative RNA-based diagnostics, likely using emerging single cell and in situ sequencing methods. In addition, we must develop comprehensive maps using preclinical models that make it possible to predict drug sensitivity (and genetic vulnerabilities) given the molecular features of the tumor. This will require expanding our repertoire of cell models (to include organoids, short-term primary tumor cultures and models that combine tumor cells and immune cells), developing more sophisticated read-outs of drug action beyond viability, and also developing new insights into the influence of the tumor microenvironment on mediating cell survival and drug resistance. The present proposal aims to address these challenges over the 7 years ahead by developing new methods and datasets that will accelerate cancer precision medicine research throughout the research community.